Estimator shortcuts

[rockstar.esq]

I was recently visiting a small project that my company had just completed.  As I was looking around, I noticed that there was a lot of unused material sitting around.  It became obvious to me that our field staff had overestimated the material they needed on every order they had placed.  Since it was a "little" job, the oversight folks weren't really paying attention to the orders.

All totaled, the guys ordered roughly 60% more material than they needed.  Virtually everything that was left over was in an unopened state.  Our PM tells me that the guys "don't have time to estimate" every little thing they need so they over-order to ensure that they can make production goals.

Setting aside the fact that I provide a material count that literally defines every single part necessary for the entire job, I considered their position.  It occurred to me that hedging a guess in order to "save time" is a real problem.  The reason I can see it's 60% more than I estimated isn't because I memorized the quantities of every part on a job I bid six months ago.  It's because I can (roughly) estimate what was needed to do the finished job I was looking at.  It dawned on me that I'm using shortcuts that allow me to quickly estimate what I'm looking at.  I figured they might prove helpful to others.

First and foremost, most of us struggle with doing math in our heads.  We've all heard people refer to "nice round numbers" when they're trying to simplify a calculation.  The problem with this tactic is that the more "round" the numbers are, the further they get from a useful answer.  The key is to mitigate mistakes by accounting for the rounding effect of your math.

Lets start with something that seems pretty obvious.  The relative precision of your calculations should be tailored to the increments of your work.  For example, thinwall conduit is 10' long and it's typically sold in 100' bundles.  Knowing that, we'd want to consider the path the conduit would take in 10' increments.  If say, your calculated run comes to 30', you know you could have three such runs per bundle.  The remaining 10' of conduit in that bundle gives you a 10% waste factor to make up for errors due to the rough measuring you're doing.  It's pretty critical to keep this straight.  In the above example, I have a 10% surplus in my calculations. 

Lets apply this to area now.  Commercial carpet is often sold by the square yard which is equal to nine square feet.  Division is easier in multiples of ten.  Lets say you have 290 square feet.  Divide that by ten and you get 29.  We know that 9 out of 10 is 90% so we have underestimated by 10%.  Shift the decimal point one to the left and you get 2.9 which rounds to 3 because we're only dealing with whole square yards.  Add 29 to three and you get 32 square yards.  If you divided 290 by 9 you'd get 32.22, so we're really close to the correct answer doing simple math in our head.

Lets take it to volume next.  Commercial concrete is usually sold by the cubic yard.  There are 27 cubic feet in a cubic yard.  It's easier to do the math using 30 instead of 27.  Notice how 10% of 30 is three which is the difference between 30 and 27?  That means we can use the same tactic as above on the area.  

Before we get to that, we should tackle another aspect of volume calculations.  In most concrete calculations, there's a definitive shape that can be visualized as an ongoing assembly.  For example a column is a prism defined by the area of the cut section multiplied by it's length.  A 12" square column is using 1 cubic foot of concrete for every foot of height.  This applies to flat stuff like floors as well.  Every square foot of floor is a constant assembly too.  

Lets say we're trying to calculate the necessary concrete to pour a 4" thick slab on a rectangular area that's 20' by 30'.  Right off the top  we know that the area is 600 square feet but we're going to have an issue with thickness in inches.  Four inches represents one third of a foot.  Another way to say this, is that we can divide our area by 3 to get volume in cubic feet.  600 divided by three is 200.  200 divided by 30 is hard enough that we can safely say it's "six and change" somewhere above 6 but below 7.  Put it down as 6.5, and repeat the 10% trick which tacks on .7 to our tally for a total of 7.2 Cubic yards.  200 Cubic feet divided by 27 gets us 7.4 Cubic yards, so here again we're really close using math we can do in our head.

Now I'm not suggesting that these sorts of rough estimations are sufficient to bid work or to place a super-critical material order.  Errors are magnitude dependent as well.  In other words, the answers will be closer to correct when you're dealing with smaller figures.  

The higher the stakes, the more precise you should be.  That being said, it can be really helpful to do a quick "gut check" on what you're seeing.  Even top-of-the-line estimating programs are susceptible to calculation mistakes.  I've caught quite a few this way.

 

 

[Glenn]

Thank you Rockstar. This is very valuable to get close. This close measure will confirm if the actual calculated measure is correct.  If they are not close, then there is a problem, which means more guesstimates and more calculations are needed.

There is always a need to go a little heavy when ordering. Guesstimates and calculations presume that things are uniform and square.  With concrete for example, the forms may not be straight, the bottom may not be level, and so on. An out of level bottom may have a thickness change of an inch or inches. This changes the volume of concrete needed. The cost of bringing an extra yard of concrete to the job site can be very expensive, not to mention getting it to blend with the concrete that has already been poured.

In another post on the site there is a detailed explanation on how to order and use steel stock. The stock is usually in 20 foot sections, so cut the stock in the following even increments, 10, 5, 4, 2, and one foot lengths. A 10 becomes 2 ea. 5 foot lengths, a 5 becomes a 3 and a 2 foot length, and so on. If you develop your designs to use even lengths, you should have little or no scrap left over. Always use the shortest piece of stock that will get the job done, for instance use a 1 foot piece of stock and do not cut one foot from a 10 foot length.

[rockstar.esq]

Glenn,

Those are really good points.  As I was reading your response it occurred to me that my examples were for materials that typically require a low percentage of waste.  Wood for trim or flooring will have a much higher percentage of waste because most installations will require straight, unblemished, and properly oriented pieces.  For example, nobody wants to look at a joint in a straight section of trim.  Flooring/decking can be tricky if the material is prone to warping.  Sometimes the waste needs to be upwards of 20% to ensure that there will be sufficient usable material for the installation.

For blacksmiths, salvaged steel can cause similar problems.  A piece of "mystery steel" that doesn't forge, bend, punch, or weld like the rest might jeopardize the entire job.  Working along those lines, if using performance spec steel, it might be better to buy a large batch for the entire thing that to risk a metallurgical difference by buying out of different production lots.

[arkie]

Glenn, the 10, 5, 4, 2 and 1 foot cuts are something I need to keep in mind.  Thanks for posting that!

[BIGGUNDOCTOR]

The gunsmith I worked for was in construction for 22 years. He did the ordering for an apartment building and he said that all of the trash from that job fit in the back of a pickup.  He really had to keep an eye on the guys to keep them from grabbing a random board and cutting it. He had it planned so when all of the pieces were cut there was nothing left over. I believe that was in the 60's.